I Took 10 DNA Tests and Compared Them | Which One Should You Take?

Which DNA test is best for African Americans? https://www.youtube.com/watch?v=-_1tgdRfHXA

You are still missing one MAJOR fact. If you buy an ancestry test, you can upload it to my heritage for free. If you buy a my heritage test, you CANNOT upload it to ancestry, therefore, if you are interested in genealogy, ONLY go for ancestry, and from there, you can upload it to my heritage and livingDNA for no cost whatsoever.

I work in forensic genetics and do those analyses every day for unidentified bodies or kinship analysis for living people. Those companies are scamming people as you can see when you compare them. If they would use scientifically proven methods, the results would match exactly. This is why we are doing interlaboratory test with other facilities around the world all getting the same samples and the results have to match 100% to be certified doing this. There is some leeway when using Y-chromosomal DNA when interpreting the results, depending on if you are using STRs or STRs plus SNPs, but they should not contradict but rather be more specific. Same with mtDNA, if you sequenced the whole genome or just the D-Loop. Doing ancestral analysis with autosomal DNA is bogus. There are some alleles you will find more often in people from region A than from region B, but even if we are talking percentages of about 90% or 99%, will still give you a 10% or 1% error rate. If your database is 20 million people, that would be 2 million or 200000 false results. Do not let high percentages fool you. A paternity test usually gives you a 99.999999% match if you are the father using just around 20 DNA systems, as long as no close relatives are also possible contenders. 99% is not really saying anything. Even if you combine all the thousands of studies done for different ethnicities and their autosomal DNA alleles, you cannot use that data to say with some certainty that this person with those alleles is from that ethnic group, let alone splitting up this group and give you even more detailed information. The Viking score is, of course, also bogus. As are differences between France and Germany, Norway and Sweden and Denmark etc. You could also roll a dice and be happy with the results without paying money for someone telling you random stuff. Now, the only real way to get ancestry information in the way you want, going back generations, is mtDNA and y-chromosomal DNA. And for those, the results have to match because they should be all based on the revised Cambridge Reference Sequence (rCRS) and use the same databases (EMPOP, Mitomap, etc.) which use the same mutations to get to the mitochondrial haplogroups. If they do not give you your data (usually given as deviations from the rCRS) do not bother with them. That way you can just check their work for yourself using the same free databases. Same with Y-chromosomal DNA. We all use the YHRD database. If they do not give you your Y-STR profile, do not bother with them. Actually, they should always provide the data they used to get to the results, so you can check it. As you can see, there is no scientific reason the results should vary so much. Or any reason they should be able to get any ancestral results from autosomal DNA alone that would reliably work. All this "10% Roman, 71% Viking, 3% Inuit"-stuff is just made up. The underlying scientific data is used in a way it was never supposed to be used and thus any references to those articles just want to lull you into a false narrative. And don't get me started on those health related results... they use results, sometimes contradicting each other, from thousands of articles, lump them together and build their own "database" for certain ailments. Some are rather easy and only rely on one or two genes and are thus easy to predict and accurate. Others are under the influence of dozens or hundreds of genes, not yet fully understood. Some paper might tell you SNP AB in Gene X will get you a 70% chance of that disease, other tell you SNP CD in Gene Y will reduce your risk by 10-40% and SNP EF in Gene Z gets you a 20% chance. Not one study will have looked at those three combined, but the company algorithm will lump those chances together as if they were, despite them probably interacting and ignoring the other 84 from other publications and the 93 not yet discovered. But they give you a 50% chance of dying of a cardiac disease anyway. Yet, even 10% or 90% mean nothing in this context. As we could see, the rheumatoid arthritis was right (with a prevalence of 1-2% in the population), the insomnia was not (with a prevalence of up to 30% in the population it was a nice try). Except the few that have a distinct genetic pathway linked to a small number of polymorphisms, all others are just an elaborate guess. Genetic dispositions are often as accurate as horoscopes... Instead of paying good money to these companies, just pay a certified and accredited lab to analyse your DNA and use the free databases yourself to get the real results. They are not as flashy as those you get from them, but feel free to embellish them yourself. Use a dice or toss a coin to determine your Viking Score...it's about as accurate that way. Or just say you are 70% Icelandic Viking if that makes you feel better. But it would be nice, if you put your DNA in a public database so we can get your murderous uncle behind bars. He has always made you feel uneasy anyway and now you know why^^

Ancestry DNA helped me find my brother that I never even knew I had or existed. I’m so thankful for them ❤

Thank you for being the tester of many & showing side-by-side results of each. Appreciate the work involved to share & teach.

I would be interested to see what happens when you send 2 or 3 samples to the same provider. Before you can talk about accuracy, it is important to know if the results are even reproducible.

The bad thing about MyHeritage is that being from Latin America does not differentiate between your percentage of indigenous and mestizo, but mixes everything as Mesoamerican and Andean, also in America there are many indigenous peoples and it seems that they only take into account the most famous indigenous people such as the Incas, Mayans and Aztecs. Being Chilean, I probably have Mapuche or aymara blood (indigenous from Chile) and I would like the companies to differentiate between these peoples. I uploaded my results in case you want to see them.

I have just binged on 8 or so of your youtube videos (mostly historical ones, but this one caught my eye). You have a gift for explaining complex topics in an "easy to digest" way. You also have a great verbal cadence which is hard to master...not too fast, not too slow with intonation in check. Keep it up!

I like narrator's clear voice, enunciation , tone & speaking speed as it is easy to understand, along with the charts -- very interesting and informative.

I'm a researcher in Genetics. One issue you didn't cover was how easy it is to get your raw data. I think most of them allow you to request the results, but I remember that several years ago, HELIX didn't provide a way to get your raw genotype data. They wanted to keep your data and sell you lots of extra tests. I would not buy a test that didn't give me full access to my data. Also, if I have the data, I don't need to pay someone to compute my Polygenic Risk Scores (PRS) since I can do that myself. More importantly, you didn't mention that a PRS may not be accurate if someone is of an ethnicity not well represented in the reference study (which is particularly a concern for people with non-european ancestry). If you do get the raw data from multiple tests, it would be interesting to compare the data. There are no absolute gold standards here, and all the tests with have some (generally very few, but some) errors. One can look at Kappa statistics between the various tests. Personally, I'd like to get whole genome sequence for myself, but from blood, not a spit or swab kit. In our research, we've seen that the spit kits are less reliable than DNA taken from blood draws.

A major but valuable investigation would be taking all nine existing tests once more, to see if some results deviate significantly between rounds!

Can we just pause to appreciate the remarkable diction and clarity of speech used to convey the information? It's abundantly clear that you are intentionally making an effort to enhance the clarity of speech, and some of us REALLY appreciate it!

You did an excellent job in explaining all the various tests, especially using your great charts. Thanks!

You forgot to mention that CRI Genetics goes back farther in your dna time line. They also tell you when in time that specific dna entered into your heritage. I’ve taken several DNA tests and this is why I liked CRI the best.

Starting in 1990, it took 13 years and billions of dollars to sequence the first human genome, and now they can do it for less than $1,000. That's utterly astonishing. (And you could argue it took 32 years depending on your definition of "fully complete genome")

Good review. I have used Ancestry and they do periodically update your information, as they get more data. Mine has changed quite a bit over about 5 years, so another thing to consider, as the first results were more broad brush and now, they seem to be homing in on distinct areas.

20 YEAR OLD COLD CASE. Detectives in Jersey contacted me! I guess I share close DNA to a murderer. I allowed them to search through my DNA relatives trying to locate this person.

One that is similar to Nebula but much cheaper is "ADNtro". It also shows the scientific papers that they got their info from, their reliability, the risk, etc. It's really nice for all of this health analysis while also including much more such as genealogy, and other miscellaneous stuff.

I originally tested with 23&Me and have uploaded to My Heritage, FTDNA, GedMatch and Living DNA. The results were pretty similar for ethnicity. I've gotten the best matches on My Heritage and GedMatch, with Living DNA being a bang on match with my tree research into my UK ancestors. Via My Heritage we discovered our Swedish ancestors actually originated in Finland.

Did my heritage a couple of years ago and it’s interesting to see how the ethnicity results have ever so slightly changed over the last couple of years to be slightly more detailed. Guessing that’s as they have more and more people from different regions signing up.